Androgen deprivation therapy (ADT) is standard palliative therapy for men with metastatic prostate cancer. Over the past two decades, it has also been increasingly used with clinically localized disease as primary monotherapy, or as salvage therapy for recurrent disease following curative-intent surgery or radiotherapy. An estimated 600,000 men with prostate cancer receive ADT each year for these different indications in the U.S. However, numerous observational studies report increased risk of serious adverse events due to ADT use, primarily bone fracture, diabetes, heart attack, and cardiac death. The impact of these adverse events on prostate cancer survivors is significant, given the prevalence of ADT use and the long survival period for most men diagnosed with prostate cancer. Additionally, there is uncertainty about the optimal application of ADT - continuously, or intermittently with off- treatment periods if a patient's serum prostate-specific-antigen (PSA) is less than 4ng/ml (reference level). Intermittent ADT has emerged as a viable alternative to continuous ADT based on recent randomized clinical trials from outside the U.S. showing that intermittent ADT may improve quality of life while retaining a survival rate equivalent to that of continuous ADT. However, it is unclear whether this evidence has altered U.S. practice regarding ADT, and no studies have examined how ADT is administered in U.S. general clinical practice. Specifically, there is little information about adoption of intermittent ADT versus continuous ADT, or whether intermittent ADT is associated with reduced risk of serious adverse events. In this proposal we will investigate the dissemination and outcomes of intermittent vs. continuous ADT in a large, population-based retrospective cohort of approximately 54,000 men in a linked cancer registry-Medicare claims database. We will first stratify ADT-treated prostate cancer patients into two groups by timing of ADT (primary monotherapy, and salvage therapy following surgery or radiotherapy). In each group, we will examine key patient and health care provider factors associated with intermittent ADT use in prostate cancer patients aged 65 or older diagnosed between 2004- 2007. Next, we will compare risks of serious adverse events between intermittent and continuous ADT in each sub-group using state-of-the-art statistical methods to account for biases in observational data. The proposed study is significant in filling an important knowledge gap regarding adoption and impact of intermittent ADT in U.S. general practice. Findings from the proposed study will generate new hypotheses regarding the comparative effectiveness of intermittent and continuous ADT and will inform subsequent research to help guide ADT use in prostate cancer.